Yield gaps in Dutch arable farming systems: Analysis at crop and crop rotation level

Abstract

Arable farming systems in the Netherlands are characterized by crop rotations in which potato, sugar beet, spring onion, winter wheat and spring barley are the most important crops. The objectives of this study were to decompose crop yield gaps within such rotations into efficiency, resource and technology yield gaps and to explain those yield gaps based on observed cropping frequencies and alternative farmers' objectives. Data from specialized Dutch arable farms between 2008 and 2012 were used. Production frontiers and efficiency yield gaps were estimated using the stochastic frontier framework. The resource yield gap was quantified through the estimation of highest farmers' yields (YHF, average across farms with actual yields above the 90th percentile). Crop model simulations and variety trials were compiled to assess climatic potential yields (Yp) and technology yield gaps. The contribution of crop area shares and farmers' objectives to actual yields were assessed using regression analysis and based on five different farm level indicators (N production, energy production, gross margin, nitrogen-use efficiency and labour use), respectively.The average yield gap per crop (as percentage of Yp which is given in parentheses) was: 29.2% (of 72.6t ha−1) for ware potato, 39.7% (of 71.6t ha−1) for starch potato, 26.4% (of 107.1t ha−1) for sugar beet, 32.3% (of 88.3t ha−1) for spring onion, 25.2% (of 12.3t ha−1) for winter wheat and 37.5% (of 10.4t ha−1) for spring barley. The efficiency yield gap ranged between 6.6% (starch potato) and 18.1% (spring onion) of Yp. The resource yield gap was lower than 10% of Yp for all the crops and the technology yield gap ranged between 7.1% (ware potato) and 30.7% of Yp (starch potato). There were statistically significant effects of potato (positive quadratic) and onion (positive) area shares on ware potato, sugar beet and winter wheat yields, of sugar beet area share (positive quadratic) on winter wheat yield and of cereal area share (negative) on sugar beet and winter wheat yields. Farmers' objectives explain part of the variability observed in crop yields which were 7–24%, 13–24% and 12–32% lower than YHF, respectively, for gross margin maximising, labour minimising and N use efficiency maximising farms. In addition, there was a significant positive relationship between gross margin and the yield of ware potato, sugar beet and winter wheat. By contrast, no significant relationships were found between crop yields and NUE or labour use.We conclude that most of the yield gap is explained by the efficiency yield gap for ware potato and spring onion and by both the efficiency and technology yield gaps for sugar beet and cereals. The resource yield gap explains most of the yield gap of seed potato, and the technology yield gap of starch potato. The results regarding the effects of cropping frequency and crop rotations to crop yields are not very conclusive which suggest that agronomic principles become less evident at ‘systems level’ given the number of interacting factors at crop rotation level. Finally, although N and energy production are lower for gross margin maximising farms, most crop yields are not significantly different between farms with the highest N and energy production compared to farms performing best on economic (gross margin) objectives.